Objective: To evaluate thresholds of various biomarkers for defining excess insulin activity to recognize congenital hyperinsulinism.
Study design: This was a retrospective chart review of diagnostic fasting tests in children with ketotic hypoglycemia (n = 30) and genetically/pathology confirmed congenital hyperinsulinism (n = 28). Sensitivity and specificity for congenital hyperinsulinism were determined for plasma insulin, β-hydroxybutyrate, free fatty acids (FFA), C-peptide, insulin-like growth factor binding protein-1 (IGFBP-1), and the glycemic response to glucagon (through the glucagon stimulation test [GST]) at the time of hypoglycemia.
Results: Only 23 of the 28 subjects with congenital hyperinsulinism had detectable insulin (median, 6.7 μIU/mL), and insulin was undetectable in all subjects with ketotic hypoglycemia. Compared with ketotic hypoglycemia, subjects with congenital hyperinsulinism had higher GST values (57 vs 13 mg/dL; ΔGST ≥30 mg/dL in 24 of 27 subjects with congenital hyperinsulinism vs 0 of 30 subjects with ketotic hypoglycemia) and C-peptide levels (1.55 vs 0.11 ng/mL), with lower levels of FFA (0.82 vs 2.51 mM) and IGFBP-1 (59.5 vs 634 ng/mL). At the time of hypoglycemia, the upper limits of β-hydroxybutyrate and FFA in subjects with congenital hyperinsulinism were higher than reported previously (β-hydroxybutyrate <1.8 mM and FFA <1.7 mM), providing the best sensitivity for congenital hyperinsulinism vs ketotic hypoglycemia. A C-peptide level ≥0.5 ng/mL was 89% sensitive and 100% specific, and an IGFBP-1 level ≤110 ng/mL was 85% sensitive and 96.6% specific.
Conclusion: Because low or undetectable insulin level during hypoglycemia does not exclude the diagnosis of hyperinsulinism, C-peptide and IGFBP-1 may inform the diagnosis of congenital hyperinsulinism. In this group of children with well-defined congenital hyperinsulinism, thresholds for "suppressed" β-hydroxybutyrate and FFA are higher than previously reported levels.
Copyright © 2016 Elsevier Inc. All rights reserved.